Description
If you want to know how pitch moves across a recording, which stretch is a vowel, or where a speaker suddenly raised their voice, listening only gives an impression, not numbers.
Praat is free software from the phonetic sciences group at the University of Amsterdam. Open a recording and you get the waveform and spectrogram, with pitch, loudness and formants measured directly. Linguists and speech therapists have used it for a long time, and plenty of the speech data in published papers comes from it.
The interface is a plain, old-school control window and assumes some phonetics background. The upside is that it supports scripting, so a measurement repeated hundreds of times can be written once and run in bulk.
Record and play back: Record directly or open audio files and replay selected segments.
Spectrogram and waveform: Vowels, consonants and pitch changes are easy to tell apart on the display.
Pitch, intensity and formants: Numeric values along the timeline, exportable for statistics.
Speech annotation: Mark syllables, words and intonation on layered tiers and save them as annotation files.
Change pitch and speed: Adjust pitch or tempo without wrecking the recording, for speech experiments.
Script automation: Batch-process large sets of recordings with the built-in scripting language so experiments can be repeated.
Praat is free software from the phonetic sciences group at the University of Amsterdam. Open a recording and you get the waveform and spectrogram, with pitch, loudness and formants measured directly. Linguists and speech therapists have used it for a long time, and plenty of the speech data in published papers comes from it.
The interface is a plain, old-school control window and assumes some phonetics background. The upside is that it supports scripting, so a measurement repeated hundreds of times can be written once and run in bulk.
Features
Record and play back: Record directly or open audio files and replay selected segments.
Spectrogram and waveform: Vowels, consonants and pitch changes are easy to tell apart on the display.
Pitch, intensity and formants: Numeric values along the timeline, exportable for statistics.
Speech annotation: Mark syllables, words and intonation on layered tiers and save them as annotation files.
Change pitch and speed: Adjust pitch or tempo without wrecking the recording, for speech experiments.
Script automation: Batch-process large sets of recordings with the built-in scripting language so experiments can be repeated.
